Lookback: Delivery Percentage, The Conviction Filter Sitting In Plain Sight

Traded volume tells you how many shares changed hands. Delivery percentage tells you how many of those buyers actually intended to own the stock past the closing bell. The gap between the two is one of the most underused conviction filters available to an Indian retail investor, and it costs nothing to read.

For a long time I treated a volume spike as a signal in itself. A stock lights up on high volume, the chart breaks out, and the instinct is to read the crowd as committed. Then the move fades within a session or two and the breakout turns out to be a trap. The piece that was missing sat right there on the exchange's own daily file: of all that frantic volume, how much was delivery, real ownership settling into demat accounts, and how much was intraday churn that netted flat by the close. Once I started reading the two numbers together, a lot of false signals resolved into what they always were.

What the number is, and what it is not

Every equity trade on the Indian exchanges falls into one of two buckets by the end of the day. Either the buyer takes delivery, the shares move into a demat account and ownership genuinely changes, or the position is squared off intraday and no delivery happens. The exchange publishes, per stock per day, the quantity that resulted in delivery as a share of the total traded quantity. That ratio is the delivery percentage.

The mental model that makes it useful is simple. Intraday churn is the footprint of traders. Delivery is the footprint of owners. A day can produce enormous volume that is almost entirely churn, a lot of hands passing the same shares around without anyone deciding to hold. Another day can produce quieter volume where a high proportion settles as delivery, fewer hands, but hands that chose to own. The first day looks exciting and means little. The second day looks dull and means a great deal.

Delivery percentage is not a timing tool and it is not a price target. It is a conviction filter, a way of asking whether the money behind a given move intends to stay.

The four readings that matter

The framework comes alive when you cross delivery percentage against price and volume, because the same delivery reading means opposite things in different contexts.

Price up, volume up, delivery percentage high. This is the cleanest bullish confirmation the number offers. The advance is not a crowd of day traders passing shares around; a meaningful share of the buying is settling into ownership. Accumulation on rising price with rising delivery is the footprint of investors building positions they intend to hold, and moves built on it tend to have staying power.

Price up, volume up, delivery percentage low. This is the trap the raw volume spike hides. The rally is running on churn, not ownership. The buyers are mostly intraday, squaring off by the close, and there is no committed holder underneath to defend the level. Breakouts on high volume but low delivery are the ones most likely to fail, and reading the delivery number is how you separate them from the real thing before you commit.

Price down, volume up, delivery percentage high. Read this carefully, because it cuts both ways. Heavy delivery-based selling into a falling price is genuine distribution, real owners handing over stock, and that is a warning. But heavy delivery on weakness can also mark the point where strong hands are absorbing panic selling and taking delivery of stock the frightened are dumping. Which one it is depends on where the move sits in the larger structure, which is why this reading is never used alone.

Price flat, volume low, delivery percentage quietly rising over many sessions. This is the least visible and often the most valuable. A stock going nowhere on unremarkable volume, but with delivery creeping up session after session, is frequently a stock being quietly accumulated by patient money before any move shows on the chart. The tape is boring on purpose. The delivery number is where the quiet accumulation leaves its fingerprint.

The disciplines that keep it honest

Three rules keep this from becoming another number you fool yourself with.

Read the trend in delivery, not the single day. One high-delivery session inside a noisy stretch is not accumulation. A rising delivery trend across many sessions, especially against a flat or basing price, is. As with most conviction signals, the persistence carries the information, not the individual print.

Compare a stock to its own baseline, never to another stock. Delivery percentage runs structurally different across names. A heavy, widely-held large cap and a tightly-held mid cap live in different delivery bands as a matter of ownership structure, not conviction. The signal is a stock breaking meaningfully above its own recent delivery range, not clearing some universal threshold. Cross-stock comparisons of the raw number are close to meaningless.

Pair it with price structure. Delivery percentage answers who is behind the move, not where the move goes. High-conviction accumulation in a stock that is structurally broken can still bleed money for a long time; committed owners can be wrong. The number filters conviction, the chart and the fundamentals decide direction. Use them together or not at all.

Where it breaks

The honest limitation is that delivery percentage is a lagging, ownership-side read, and it says nothing about why the owners are buying. It cannot distinguish informed accumulation from misguided conviction, and it can be distorted in names where a single large holder's block moves dominate the print. In thinly-traded stocks the ratio gets noisy enough to be unusable. Treat it as a filter on the quality of a move you already have other reasons to be interested in, never as a standalone reason to act.

Building it into a routine

The way to make delivery percentage useful is to fold it into a screening routine rather than glancing at it after the fact. Start from a shortlist you already care about, names with a structural or fundamental reason to be on your radar, and pull each one's delivery percentage against its own trailing range rather than against the market. The question is never whether a stock's delivery is high in absolute terms; it is whether the stock is printing delivery meaningfully above where it normally sits.

From there the four readings do the sorting. A shortlisted name breaking higher on rising delivery moves to the top of the queue, because the move is being confirmed by ownership. The same breakout on collapsing delivery drops down the queue or off it, flagged as a probable churn-driven trap regardless of how clean the chart looks. A basing name whose delivery is quietly climbing over weeks earns a watch even though nothing has happened on price, because that quiet climb is often the tell before the move. And a name falling on heavy delivery gets a hard second look at its structure, because genuine owners handing over stock is a different warning from a thin, low-delivery drift lower.

The routine also protects you from the number's biggest weakness, which is that it flatters block-heavy days. A single large negotiated block can spike a day's delivery percentage without telling you anything about broad conviction, so read the ratio across a run of ordinary sessions and treat any single spike with suspicion until the trend confirms it. Held as a filter on a shortlist, benchmarked to each stock's own history, and read across sessions rather than in a single print, delivery percentage quietly upgrades the quality of every other signal you already use. It will not find ideas for you. It will tell you which of your ideas the owners agree with.

Verdict, what BazaarBaazi thinks

Delivery percentage is the closest thing the Indian retail investor gets to a free conviction filter, published daily by the exchange, ignored by most, and quietly decisive when read against price and volume. It will not tell you where a stock is going. It will tell you whether the people buying it mean to keep it, and that alone is enough to keep you out of a large share of the volume-spike traps that catch the crowd. Stance is neutral by design, because this is a lens on conviction, not a directional call. Read the trend, benchmark a stock against itself, and let the number do the one job it does well: separating owners from churn. Treat it as one input among several, never the whole thesis, and it will earn its place on the desk. The exchange hands you this filter for free every single trading day. The only cost is the discipline to read it in context, and the patience to wait for the trend rather than the print. Used that way, it quietly keeps you on the right side of conviction more often than not.

---

Disclosure: prepared using BazaarBaazi's editorial AI tooling, with research validation, fact-checking, and final editorial sign-off by Aditya Sharma. BazaarBaazi is YMYL finance content; every falsifiable specific is primary-source verified or stripped. This is an analytical framework, not investment advice.